Forcecontrolled walls and their application for shear tester

Outline • Material properties • Jenike shear tester • Basic idea • Experimental set-up

Material properties • Commonly used DEM contact models are spring-dashpot Material properties for Hertz

Method Determination of the coefficient of friction by means of a comparison of experimental

Jenike shear tester Shear cell of the Jenike shear tester (from Schulze D. ,

Experimental set-up Requirements for the numerical simulation: • Motion of one ring with constant

Numerical simulation • The shear cell consists of a lid and two rings •

What LIGGGHTS already can. . • A triangulated mesh can be moved and rotated

Force controller PID-controller • with anti-windup mechanism • maximum velocity limiter • This approach

Numerical set-up August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling |

Check the servo-wall σz (Pa) Normal stress applied by the servo-wall These results are

Results Example for glass beads with r = 0. 25 mm Ratio shear stress

Conclusions • A force controlled ‘servo-wall’ was implemented into the framework of LIGGGHTS •

Thank you. Questions? Johannes Kepler University Linz CD Laboratory on Particlulate Flow Modelling Contact:

Slides: 14

Download presentation

Force-controlled walls and their application for shear tester simulations LAMMPS workshop on August 8, 2013 Andreas Aigner <andreas. aigner@jku. at> CD Laboratory on Particulate Flow Modelling Johannes Kepler University | Linz | Austria

Outline • Material properties • Jenike shear tester • Basic idea • Experimental set-up • Numerical model • What LIGGGHTS already can… • Force controller • Comparison of experimental and numerical results • Conclusions August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling | www. jku. at/pfm 2

Material properties • Commonly used DEM contact models are spring-dashpot Material properties for Hertz model Radius [R] Mass [m] Young‘s modulus [E] Poisson‘s ratio [ν] • Coefficient of restitution Coefficient of friction [μr] • August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling | www. jku. at/pfm 3

Method Determination of the coefficient of friction by means of a comparison of experimental and simulation results of a simplified Jenike shear tester Average of several experimental runs Numerical results for varying coefficient of friction τxz/σz Compare of the ‚steady-state‘ flow Time (s) August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling | www. jku. at/pfm Time (s) 4

Jenike shear tester Shear cell of the Jenike shear tester (from Schulze D. , Flow Properties of Powders and Bulk Solids, 2011) August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling | www. jku. at/pfm 5

Experimental set-up Requirements for the numerical simulation: • Motion of one ring with constant velocity • Determination of the force acting on the rings Scatch of the simlified Jenike shear • A constant normal force tester used for experiments. has to be applied to the particles August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling | www. jku. at/pfm 6

Numerical simulation • The shear cell consists of a lid and two rings • Therefore triangulated meshes are imported August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling | www. jku. at/pfm 7

What LIGGGHTS already can. . • A triangulated mesh can be moved and rotated with variable or constant velocity and angular velocity, respectively. • All forces and torques acting on one geometry are calculated and accessible in the input script fix cad 1 all mesh/surface/stress file mesh/upper. Cyl. stl type 2 scale 0. 001 move 0. 0. 0. com 0. 0. 0. variable Fx. Cad 1 variable Fy. Cad 1 … equal f_cad 1[1] equal f_cad 1[2] August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling | www. jku. at/pfm 8

Force controller PID-controller • with anti-windup mechanism • maximum velocity limiter • This approach can also be used to control the torque by replacing force and velocity by torque and angular velocity. August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling | www. jku. at/pfm 9

Numerical set-up August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling | www. jku. at/pfm 10

Check the servo-wall σz (Pa) Normal stress applied by the servo-wall These results are achieved with a pure proportional controller: • The normal stress is kept constant • Only a small overshoot at the first particle-wall contact Time (s) August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling | www. jku. at/pfm 11

Results Example for glass beads with r = 0. 25 mm Ratio shear stress / normal stress versus time Time (s) August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling | www. jku. at/pfm 12

Conclusions • A force controlled ‘servo-wall’ was implemented into the framework of LIGGGHTS • This new wall type provides a constant normal stress for the simulation of a simplified Jenike shear tester. • By comparing experimental and numerical results the coefficient of friction is determined for individual granular material. August 5, 2013 | Christian Doppler Laboratory on Particulate Flow Modelling | www. jku. at/pfm 13

Thank you. Questions? Johannes Kepler University Linz CD Laboratory on Particlulate Flow Modelling Contact: andreas. aigner@jku. at